On Tue, May 24, 2011 at 01:10, Tanya Brokhman wrote: >> >+ Â Âcase USB_REQ_GET_STATUS: >> >+ Â Â Â Â Â Âif (!gadget_is_superspeed(gadget)) >> >+ Â Â Â Â Â Â Â Â Â Âgoto unknown; >> >+ Â Â Â Â Â Âif (ctrl->bRequestType != (USB_DIR_IN | USB_RECIP_INTERFACE)) >> >+ Â Â Â Â Â Â Â Â Â Âgoto unknown; >> >+ Â Â Â Â Â Âvalue = 2; Â Â Â/* This is the length of the get_status >> >+ Â Â Â Â Â Â*((__le16 *)req->buf) = 0; >> >> Mike please correct me if I'm wrong bug this looks like a case for >> put_unaligned_le16(). >> Is someone actually using gadget support on blackfin? I'm asking >> because >> config_buf() (same file, upstream) is using req->buf to build the >> descriptors in place and one element is le16 which should be affected. > > Mike answered that you're right in your observation. I'm not familiar with > blackfin. Could you please elaborate on this? I understand that I need to > use put_unaligned_le16(), will do, but I would like to better understand why > and if there is a way to test this so that blackfin won't be broken. not all arches support unaligned accesses. or they do, but it's done via (non-trivial) exception processing in software. req->buf is of type void* and so presumably is not guaranteed to be aligned on a 2 byte boundary. -mike -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
